In 1981 one of us (Cherry) first attempted to identify spatial and temporal patterning in the human colonization of the Mediterranean islands. Since the 1980s, slowly accumulating evidence has suggested that the Mediterranean islands were sporadically inhabited by hunter-gatherers during the Late Pleistocene and Early Holocene. Here we seek to establish whether or not these data exhibit regularity. We suggest that evidence for Upper Palaeolithic to Mesolithic activity, tending to cluster on larger or less remote islands, indicates that while humans were clearly capable of reaching the Mediterranean islands prior to the Neolithic, their general reluctance to do so can be explained in terms of the variable environmental attractiveness of the insular Mediterranean. Tending to be relatively small, dry, and biologically depauperate, the Mediterranean islands were largely inhospitable to mobile groups preferring extensive territories with diverse and robust biotas. Sedentism only became a widely viable strategy in the insular Mediterranean with the development in the Neolithic of what we might regard as “terraforming”—that is, the introduction of cereals, pulses, and ovicaprids, all tolerant of xeric environments.